KPP OBLdepth needs horizontal smoother option about mom6 HOT 5 CLOSED

Before adding a horizontal smoother,
  please recall that because MOM6 is a C-grid model and has no null
  mode in density, it is vastly more tolerant of grid-scale
  variations in mixing rates than a B-grid model would be. Is there
  any symptom, apart from noise in the PBL depth itself, that
  indicates a need for smoothing?

    - Bob Hallberg

  On 03/18/2015 04:34 PM, Stephen Griffies wrote:


  We need to code a horizontal smoother option for KPP boundary
    layer depth. There is noise, particularly
    in the tropics. Noise is the KPP OBLdepth is well known, with
    all groups making use of KPP finding
    the need to perform a smoothing operation on the BLdepth prior
    to computing diffusivity and 
    nonlocal transport. 
  —
    Reply to this email directly or view
      it on GitHub.

from mom6.

Hi,

Good point. The noise is indeed less than B-grid, but it is there. I
doubt if it is of much significance. I only think it useful to have in
case there are places where it is a problem. I also note that ROMS
performs a smoothing for the same reason as I alluded to.

Does the bulk mixed layer show any noise? Is is a one-dim column model in
the sense that horizontal columns are independent?

Steve

On Wed, Mar 18, 2015 at 5:06 PM, Robert Hallberg [email protected]
wrote:

Before adding a horizontal smoother,
please recall that because MOM6 is a C-grid model and has no null
mode in density, it is vastly more tolerant of grid-scale
variations in mixing rates than a B-grid model would be. Is there
any symptom, apart from noise in the PBL depth itself, that
indicates a need for smoothing?

• Bob Hallberg

On 03/18/2015 04:34 PM, Stephen Griffies wrote:

We need to code a horizontal smoother option for KPP boundary
layer depth. There is noise, particularly
in the tropics. Noise is the KPP OBLdepth is well known, with
all groups making use of KPP finding
the need to perform a smoothing operation on the BLdepth prior
to computing diffusivity and
nonlocal transport.
—
Reply to this email directly or view
it on GitHub.

—
Reply to this email directly or view it on GitHub
#92 (comment)
.

Dr. Stephen M. Griffies
NOAA Geophysical Fluid Dynamics Lab
201 Forrestal Road
Princeton, NJ 08542
USA

from mom6.

Hi Steve,

    Yes, a bulk mixed layer would show noise, in that each column
  involves an independent calculation. However, since the boundary
  layer depth with a bulk mixed layer is based on an integrated
  energy budget, and not some fairly arbitrary point difference
  calculations, it is vastly less sensitive than KPP to small
  changes in details of how interpolation is done or in the exact
  values at various points.  I don't know why ROMS would need to do
  smoothing, but perhaps it is to counteract some of the twitchiness
  in the KPP calculations.

    - Bob Hallberg

   Dr. Robert Hallberg

Oceanographer and Ocean Group Head
NOAA Geophysical Fluid Dynamics Laboratory

NOAA GFDL || Phone: (609) 452-6508
Princeton University Forrestal Campus || Home: (732) 463-1274
201 Forrestal Road || Fax: (609) 987-5063
Princeton, New Jersey 08540-6649 || Email: [email protected]

  On 03/18/2015 05:16 PM, Stephen Griffies wrote:

Hi,

  Good point. The noise is indeed less than B-grid, but it is there.
  I
  doubt if it is of much significance. I only think it useful to
  have in
  case there are places where it is a problem. I also note that ROMS
  performs a smoothing for the same reason as I alluded to.

  Does the bulk mixed layer show any noise? Is is a one-dim column
  model in
  the sense that horizontal columns are independent?

  Steve

  On Wed, Mar 18, 2015 at 5:06 PM, Robert Hallberg
  <[email protected]>
  wrote:

  >
  >
  >
  >
  >
  > Before adding a horizontal smoother,
  > please recall that because MOM6 is a C-grid model and has no
  null
  > mode in density, it is vastly more tolerant of grid-scale
  > variations in mixing rates than a B-grid model would be. Is
  there
  > any symptom, apart from noise in the PBL depth itself, that
  > indicates a need for smoothing?
  >
  > - Bob Hallberg
  >
  > On 03/18/2015 04:34 PM, Stephen Griffies wrote:
  >
  >
  > We need to code a horizontal smoother option for KPP boundary
  > layer depth. There is noise, particularly
  > in the tropics. Noise is the KPP OBLdepth is well known, with
  > all groups making use of KPP finding
  > the need to perform a smoothing operation on the BLdepth
  prior
  > to computing diffusivity and
  > nonlocal transport.
  > —
  > Reply to this email directly or view
  > it on GitHub.
  >
  > —
  > Reply to this email directly or view it on GitHub
  >

#92 (comment)
> .
>

  -- 
  Dr. Stephen M. Griffies
  NOAA Geophysical Fluid Dynamics Lab
  201 Forrestal Road
  Princeton, NJ 08542
  USA
  —
    Reply to this email directly or view
      it on GitHub.

from mom6.

Bob,

All good points.

My understanding is that ROMS smooths the KPP noise for reasons that you
indicate; "twitchiness
in the KPP calculations." (That is a great phrase I hope to use in a paper
one day with you and Alistair!).

Steve

On Wed, Mar 18, 2015 at 5:42 PM, Robert Hallberg [email protected]
wrote:

Hi Steve,

Yes, a bulk mixed layer would show noise, in that each column
involves an independent calculation. However, since the boundary
layer depth with a bulk mixed layer is based on an integrated
energy budget, and not some fairly arbitrary point difference
calculations, it is vastly less sensitive than KPP to small
changes in details of how interpolation is done or in the exact
values at various points. I don't know why ROMS would need to do
smoothing, but perhaps it is to counteract some of the twitchiness
in the KPP calculations.

• Bob Hallberg

Dr. Robert Hallberg
Oceanographer and Ocean Group Head
NOAA Geophysical Fluid Dynamics Laboratory

NOAA GFDL || Phone: (609) 452-6508
Princeton University Forrestal Campus || Home: (732) 463-1274
201 Forrestal Road || Fax: (609) 987-5063
Princeton, New Jersey 08540-6649 || Email: [email protected]

On 03/18/2015 05:16 PM, Stephen Griffies wrote:

Hi,

Good point. The noise is indeed less than B-grid, but it is there.
I
doubt if it is of much significance. I only think it useful to
have in
case there are places where it is a problem. I also note that ROMS
performs a smoothing for the same reason as I alluded to.

Does the bulk mixed layer show any noise? Is is a one-dim column
model in
the sense that horizontal columns are independent?

Steve

On Wed, Mar 18, 2015 at 5:06 PM, Robert Hallberg
[email protected]
wrote:

Before adding a horizontal smoother,
please recall that because MOM6 is a C-grid model and has no
null
mode in density, it is vastly more tolerant of grid-scale
variations in mixing rates than a B-grid model would be. Is
there
any symptom, apart from noise in the PBL depth itself, that
indicates a need for smoothing?

• Bob Hallberg

On 03/18/2015 04:34 PM, Stephen Griffies wrote:

We need to code a horizontal smoother option for KPP boundary
layer depth. There is noise, particularly
in the tropics. Noise is the KPP OBLdepth is well known, with
all groups making use of KPP finding
the need to perform a smoothing operation on the BLdepth
prior
to computing diffusivity and
nonlocal transport.
—
Reply to this email directly or view
it on GitHub.

—
Reply to this email directly or view it on GitHub

<
#92 (comment)

.

Dr. Stephen M. Griffies
NOAA Geophysical Fluid Dynamics Lab
201 Forrestal Road
Princeton, NJ 08542
USA
—
Reply to this email directly or view
it on GitHub.

—
Reply to this email directly or view it on GitHub
#92 (comment)
.

Dr. Stephen M. Griffies
NOAA Geophysical Fluid Dynamics Lab
201 Forrestal Road
Princeton, NJ 08542
USA

from mom6.

I think we should close this issue. We are moving to EPBL, so will not need to refine the KPP approach. The above points arguing against adding smoothers are also good.

from mom6.